Does this skill work with both Python and Node.js?

Yes, it provides comprehensive integration patterns for both TypeScript/JavaScript using the @olakai/sdk and Python via the olakai-sdk.

What is the benefit of registering CustomDataConfigs?

Registering your custom data fields allows Olakai to use them as variables in KPI formulas, enabling advanced business metric tracking and alerting.

Can I monitor streaming LLM responses?

Absolutely. The Olakai SDK handles streaming responses automatically, sending the telemetry event once the stream is finished.

How do I handle agents that make multiple LLM calls?

The skill includes an 'Agentic Workflow' pattern that shows you how to aggregate multiple LLM calls into a single manual event for cohesive monitoring.

Olakai AI Monitoring & Observability

Name: Olakai AI Monitoring & Observability
Author: olakai-ai

byolakai-ai

Analytics & Monitoring

Integrates professional observability, custom KPIs, and business metrics tracking into existing AI agents and LLM applications.

About

The olakai-add-monitoring skill empowers developers to retrofit production-grade telemetry into AI systems with minimal code changes. By wrapping OpenAI and Anthropic clients or using manual event tracking, it captures more than just token counts—it enables the creation of business-specific KPIs, ROI tracking, and performance dashboards. This skill guides you through the entire pipeline of instrumenting code, registering custom data schemas, and establishing workflow-level monitoring for complex, multi-step agentic behaviors.

Key Features

Automatic client wrapping for OpenAI and Anthropic SDKs
Support for both TypeScript/JavaScript and Python environments
Multi-step agentic workflow aggregation and session grouping
CLI-integrated agent and workflow management
0 GitHub stars
Custom KPI configuration for business-specific performance tracking

Use Cases

Retrofitting observability into a production AI agent without rewriting logic
Monitoring performance trends and token costs across complex agentic workflows
Measuring the business ROI of LLM features through custom success metrics

About

Key Features

Automatic client wrapping for OpenAI and Anthropic SDKs
Support for both TypeScript/JavaScript and Python environments
Multi-step agentic workflow aggregation and session grouping
CLI-integrated agent and workflow management
0 GitHub stars
Custom KPI configuration for business-specific performance tracking

Use Cases

Retrofitting observability into a production AI agent without rewriting logic
Monitoring performance trends and token costs across complex agentic workflows
Measuring the business ROI of LLM features through custom success metrics